def main(args, _in, _out): parts = args.matcher_fn.split(".") module = __import__(".".join(parts[:-1]), fromlist=[""]) matcher = getattr(module, parts[-1]) with open(args.f, "r") as f: _dict = list(load_dict(f)) raw = list(load_raw(_in)) from rank.util.feature import vw_model bow = list(compute_bow(raw, _dict, matcher, args.max_dist, args.multi)) _out.write("{0}".format(vw_model(bow_string(bow), args.label)))
def main(args, _in, _out): import pandas as pd from rank.util.feature import parse_bow, vw_model with open(args.m, "r") as f: mask = list(map(lambda x: x.strip(), f.read().split("\n"))) data = [parse_bow(line) for line in map(lambda x: x.strip(), _in.read().split("\n")) if len(line) > 0 ] df = pd.DataFrame(data) filtered = df[mask] for label, line in zip(df["label"], filtered.values): _out.write("{0}\n".format(vw_model(" ".join( map(lambda x: "{0}:{1}".format(*x), zip(filtered.columns, line))), label) ))
def main(args, _in, _out): import pandas as pd from rank.util.feature import parse_bow, vw_model with open(args.m, "r") as f: mask = list(map(lambda x: x.strip(), f.read().split("\n"))) data = [ parse_bow(line) for line in map(lambda x: x.strip(), _in.read().split("\n")) if len(line) > 0 ] df = pd.DataFrame(data) filtered = df[mask] for label, line in zip(df["label"], filtered.values): _out.write("{0}\n".format( vw_model( " ".join( map(lambda x: "{0}:{1}".format(*x), zip(filtered.columns, line))), label)))